Project in steps

Step 1: Create python-app docker image and push this image to dockerhub and make the image a private image.

Step 2: Create docker-compose.yml that contains the services python-app,MySQL and phpmyadmin.

Step 3: Create other docker-compose.yml contains jwilder/nginx-proxy to expose python-app and phpmyadmin with a self signed certificates.

Step 4: Deploy the python-app,MySQL and phpmyadmin to k8s minikube cluster taking into account the persistence of MySQL data, Put Environment variables in Configmap, Acees applications from outside the minilube using the same domains and use self signed certificates as in docker-compose.

Step 5: Install gitlab and push previous code(docker-compose files and k8s files) on gitlab, and configure a pipeline with gitlab-ci file to auto deploy the stack on the docker environment and the kubernetes environment.

Step 6: Write down the whole k8s files (deployments,configmaps, secrets, services, ingreses, etc) to a helm chart, in other meaning, create a helm chart for the stack.

Step 7: Auto deploy the stack via ArgoCD using the helm chart that we wrote it down previously taking into account to prevent any changes from the server side to be applied, only changes can be done from the helm chart that is pushed on the repository.

Step 1: Create python-app docker image and push this image to dockerhub and make the image a private image.

• Write Dockerfile to build the flask application to docker image
• app.py and Dockerfile is in flask-app
• After building the image, Push it to the dockerhub registry in a private repo

Step 2: Create docker-compose.yml that contains the services python-app,MySQL and phpmyadmin.

• db.yml and flask-php.yml files are in docker-compose

Step 3: Create other docker-compose.yml contains jwilder/nginx-proxy to expose python-app and phpmyadmin with a self signed certificates.

-First, Create 2 self signed certificates:

• The first certificate is to flask-app with domain name called flask-app.local

• The second certificate is to phpmyadmin with domain name called phpmyadmin.local

• Certs are in docker-compose/nginx-proxy/certs

• But first we need to put these domains in our /etc/hosts file to map them to the proxy to route them

• Second, Create proxy file with the created ssl to be able to access our apps:

• flask app on flask-app.local domain name

• phpmyadmin on phpmyadmin.local domain name

• proxy.yml is in docker-compose

• Now we are able to access our apps on these domains

Step 4: Deploy the python-app,MySQL and phpmyadmin to k8s minikube cluster taking into account the persistence of MySQL data,put Environment variables in Configmap,acees applications from outside the minilube and use the same domains and use self signed certificates as in docker-compose.

Prerequisites

• Before we deploy our k8s files, We need to configure 2 things:

• First, As our flask image is in a private repo in dockerhub registry, We need to create a Pull image secret to enable flask-deployment.yml to pull the image from the private dockerhub registry

• We will do some steps:

• 1- Create .docker-config.json file

• Where auth attribute is base64 of dockerhub-username:dockerhub-password

• 2- Now,We can create our pull image secret

• Where .dockerconfigjson attribute is base64-encoded-contents-of-.docker/config.json file

• 3- After creating the pull image secret,We must refer it into flask-deployment.yml

• Now our flask deployment is able to pull the flask image from the private repo

• Second We need to add our minikube ip and our domain names to /etc/hosts to map these doamins to the minikube ingress controller to route them

Deploy our applications on the minikube cluster

• All k8s files are in gitlab-k8s

Acess our apps

• Now, We can access our apps on the specific domains

Step 5: Install gitlab and push previous code(docker-compose files and k8s files) on gitlab, and configure a pipeline with gitlab-ci file to auto deploy the stack on the docker environment and the kubernetes environment

Install gitlab locally with docker-compose

• We installed gitlab with a docker-compose file to presist data,configration and logs

• Acess gitlab UI

• install-gitlab.yml is in install-gitlab

Deploy docker-compose files by gitlab (.gitlab-ci.yml file)

• Before we push our docker-compose files we need to configure 3 things:

• First, Add docker login credentials in gitlab to enable docker-compose to pull the flask image from the private dockerhub registry

• Second, We need to configure and add a runner to our project to execute pipelines commans on it

• Our runner is active and running and ready to receive pipeline jobs

• Third, Configure .gitlab-ci.yml file and write the pipeline

• All docker-compose files,scripts and .gitlab-ci.yml file is on gitlab-compose

• Now,We are ready to deploy our docker-compose files

• As soon as the docker compose files and .gitlab-ci.yml file are pushed to the gitlab repo, The pipeline will start

• Pipeline ran successfully

• Now, We can access our apps from Environmnent tab

Deploy K8s files by gitlab (.gitlab-ci.yml file)

Prerequisites

• Before we deploy our k8s files, We need to configure some things:

• First, Create a dedicated service account with a dedicated permissions to use this service account to aceess minikube cluster and deploy k8s files

• Second,Put the token of the secret file in kubeconfig file to be able to access the minikube cluster

• To do that, We need first to create a secret token to the service account

• Then we need the token from this secret file and convert it to bas64

• Encode the token to base64

• Finally, Add the encoded bas64 token to kubeconfig file

• Now we have a kubeconfig file and can use it to access the minikube cluster and deploy k8s files

• Add this kubeconfig file to gitlab to use it in our pipeline

• gitlab service account files are in gitlab-serviceaccount

• We need also to configure and add a runner to our project to execute pipelines commans on it

• Our runner is active and running and ready to receive pipeline jobs

• All k8s files,scripts and .gitlab-ci.yml file is on gitlab-k8s

• Now,We are ready to deploy our k8s files

• As soon as the k8s files and .gitlab-ci.yml file are pushed to the gitlab repo, The pipeline will start

• Pipeline ran successfully

• Now, We can access our apps from Environmnent tab

Step 6: Write down the whole k8s files (deployments,configmaps, secrets, services, ingreses, etc) to a helm chart, in other meaning, create a helm chart for the stack.

• Stack Helm chart files is in stack-chart

Step 7: Auto deploy the stack via ArgoCD using the helm chart that we wrote it down previously taking into account to prevent any changes from the server side to be applied, only changes can be done from the helm chart that is pushed on the repository.

• First, we need to install ArgoCD on our minikube cluster

• Then, Froward ArgoCDd service to port 8080 to access ArgoCD UI

• Finally, Get ArgoCD password to access the UI

• After ArgoCD working, We can now apply ArgoCD-App.yml to configure and synchronize ArgoCD with Helm chart repo

• ArgoCD-App.yml file is in ArgoCD

• After doing that ArgoCD become synchronized with Helm chart repo and changes in applications happens only from Helm chart repo